use inplace=None as default in densenet

[aljkor]

Fix was implemented by replacing the usage of ReLU activation with inplace=True with the optional parameter inplace, that is by default None. This fixes issues with backward pass and calculating gradients.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/8306

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[facebook-github-bot]

Hi @aljkor!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@meta.com. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[NicolasHug]

Thanks for the PR @aljkor

Can you share more details on what this is fixing exactly?

[aljkor]

Hi,
for example if I try using shap.DeepExplainer on basic pytorch models (to explain the output of any machine learning model), I get the following error.

Output 0 of BackwardHookFunctionBackward is a view and is being modified inplace. This view was created inside a custom Function (or because an input was returned as-is) and the autograd logic to handle view+inplace would override the custom backward associated with the custom Function, leading to incorrect gradients. This behavior is forbidden. You can fix this by cloning the output of the custom Function.

[aljkor]

Here I found the same issue with a different model: shap/shap#3466

[peacefulotter]

Any news on this? This issue makes models such as efficientnetv2 unusable with the shap package (throws the error mentioned above by @aljko)